Epstein-Barr virus (EBV) is a B lymphotropic herpesvirus that usually persists benignly for the life time of the infected host, but can be predisposing for the development of certain lymphomas and carcinomas. Persistent infection in the peripheral blood of healthy carriers is quiescent, being restricted to latently infected, resting, memory B cells in which, at most, one or two latent genes are expressed. By comparison persistent infection in the mucosal lymphoid tissue appears to be more complex. There are infected naive and memory B cells, viral replication to produce infectious virus and at least one completely novel subset with a very high frequency of latently infected B cells. The virus is not randomly distributed through all of the B cells subsets. It is our hypothesis that the distribution of virus infected cells can be explained through analogy to normal B cell activation and differentiation within secondary lymphoid tissue. To test this idea we will focus on three main issues: 1. What is the nature of the novel B cell subset (CD38+CD10-IgD-) that we have found in the tonsils to be highly enriched for latently infected cells (approximately 1 in 500 are infected) and what is its relationship to classically defined memory B cells (CD38- CD10-IgD-) 2. Is EBV a mucosal specific virus i.e. is viral persistence actively maintained throughout the lymphoid tissue or quiescent everywhere except mucosal lymphoid tissue? Does the virus replicate in all lymphoid tissue or only mucosal lymphoid tissue? 3. Are virus infected cells in the lymphoid tissue able to undergo a germinal center reaction to enter the memory compartment.